© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 10-GEMMASCOPE-RES-131K
    7. 124095
    Prev
    Next
    INDEX
    Explanations

    show that if

    np_acts-logits-general · gemini-2.5-flash-lite

    The neuron activates on three-digit numeric tokens (e.g. problem or section numbers like “113,” “146,” etc.).

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_10/width_131k
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    No Configuration Found
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    (!__
    -2.22
    awić
    -1.67
    がこの
    -1.66
    OTROS
    -1.66
    TIPOS
    -1.65
     estatua
    -1.65
     demostra
    -1.65
    朥
    -1.64
     was
    -1.64
    }.
    -1.63
    POSITIVE LOGITS
     When
    2.06
     Sometimes
    2.02
     if
    1.96
     there
    1.84
     EVERY
    1.70
     the
    1.69
     ALWAYS
    1.69
    If
    1.66
     this
    1.65
     bedrock
    1.65
    Activations Density 0.066%

    No Known Activations